BJA Advance Access originally published online on October 24, 2007
British Journal of Anaesthesia 2007 99(6):898-905; doi:10.1093/bja/aem302
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Responsive Contingency Planning: a novel system for anticipated difficulty in airway management in dental abscess
Department of Anaesthesia, University Hospital Aintree, Longmoor Lane, Liverpool L9 7AL, UK
* Corresponding author. E-mail: peter.charters{at}aintree.nhs.uk
Accepted for publication August 17, 2007.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Background: This study is of a novel system for management of anticipated difficult airway (Responsive Contingency Planning). It is based on the notion that almost all problems in airway management have already been experienced, so they can be anticipated and prepared for using ‘worst case’ planning.
Methods: Anaesthetic colleagues were introduced to the new system. Thirty-two patients with dental abscess were recruited for anaesthetic airway management based around the new scheme. Data collection involved a preoperative assessment of problems specific to dental abscess, fascia-spaces involved, details of the contingency planning process, laryngoscopy grade, and comments regarding the efficacy of the new system.
Results: No problems were encountered that had not been anticipated during planning and colleagues' comments about using the system were generally favourable. Examples are highlighted to suggest how the planning may have avoided certain complications (e.g. abscess rupture) and helped in dealing with others when they occurred. If trismus (
2 cm inter-incisor distance) occurred with superficial infection alone, it always improved (to >2 cm) after induction of anaesthesia. In contrast, difficult laryngoscopy (grade 3 or 4) occurred in 6/15 cases of floor of mouth infection.
Conclusions: The system fulfilled expectations for its use at this stage of development. It can be easily updated for refinements, alternative techniques, and tailoring to any difficult airway scenario. Computerization should make it easier to use and flag-up inconsistencies. Floor of mouth infections in the presence of trismus are easily underestimated and require careful assessment.
Keywords: airway, infections; anaesthetic techniques, laryngoscopy; safety, techniques
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Departmental presentation of a local retrospective audit of anaesthesia for dental abscess had suggested inadequate airway management planning for these cases. Although a before operation formulated plan was usually documented, the general conclusion was that when problems did occur (in 5/29 cases) these always came as surprise requiring a total re-think. The problems ranged in severity from minor to very serious and included failures in fully assessing the implications of limited airway access, appreciating the implications of toxicity when problems occurred, and identifying a pointing abscess. We considered that a better alternative might be ‘Responsive Contingency Planning’. This new concept of management is based on the idea that all the possible problems that may occur will almost certainly have happened to others in the past. This means that they can be anticipated and prepared for using ‘worst case’ planning. We were also influenced by Perry1 when he stated ‘Never underestimate fascial space infection. In fact never call it "dental abscess" as the terminology will put you, the anaesthetist and the theatre staff in a wrong state of mind and lower gear of alertness. It is a fascial or cervical space infection.’ Fascial space infections can be difficult to assess clinically and may present the anaesthetist with quite complex problems for airway management.2–4
If our new system was satisfactory, we anticipated positive answers to two questions.
- Can the method predict all airway management outcomes in these patients?
- Will the method be seen as advantageous to the anaesthetists who use it?
|
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
After obtaining the approval of local research ethics committee, we recruited our anaesthetic colleagues for this study. A departmental presentation explained the study protocol with a review of tissue space anatomy (Figs 1 and 2) and clinical significance. Inclusion criteria were patients aged 16 yr and more, undergoing incision and drainage of dental abscess. Patients at risk of reflux or vomiting at induction were excluded. The study forms are reproduced as Parts A–D (Fig. 3, Tables 2–4). Colleagues would consent their patients for this study at the preoperative visit and sketch the abscess outline on the clinical diagrams before operation (Part B).
|
|
|
|
Part C (Table 2) of the form contained a list of problems known to be associated with dental abscess. At the preoperative visit, anaesthetists noted which were present and subjectively scored them 1–5 (5 being worst), prompted by their answers for Part B. To make the next stage easier to understand, Table 2 simply indicates whether a problem was present by a tick in the second column.
In our institution four management options were routinely practised:
- conventional direct laryngoscopy;
- intubation using an Aintree catheter, flexible fibrescope, and LMA-Classic®;
- awake nasal flexible fibreoptic laryngoscopy;
- awake surgical tracheostomy.
Individual matrix cells were scored: X, contra-indicated; C, caution; or I, indicated according to how the anaesthetist perceived the problem for the individual patient and their own ability with the relevant management option. Each problem had to be considered as though it was occurring in isolation from the others. Finally, each management option was then summarized according to its own worst score (when X>C>I) in the final row. (For example, in the ‘GA/LMA’ option, there are ‘X’, ‘I’, and two ‘C’ scores so the result is ‘X’, and similarly for ‘A Trach’, there are three ‘I’ and one ‘C’ scores so the result is ‘C’.) The winner is the management with the best summary score (in this case ‘AFOL’) and in the case of ties, the fewest adverse scores and a review of problem severities determines the winner.
Part D is shown in contingency tables (Tables 3 and 4) for direct laryngoscopy and awake fibreoptic intubation, the only managements used in this study. Each is outlined as a sequence of steps leading to tracheal intubation. All are subdivided into left and right sides which represent expected ‘likely problems’ for the particular management option (left) and ‘things making it worse’ particularly in relation to the identified clinical problems (right). The columns filled in with italics are representative responses because these sections of the form would normally be blank. Entering responses to ‘my solutions’ (i.e. to ‘likely problems’) and ‘plans if worst happens’ (i.e. to ‘things making it worse’) was how the anaesthetist refined plans for an individual patient. Trainees either provided their own answers or asked their consultant supervisor. The trainee and consultant would then discuss the need for supervision of the case and how nearby that supervision should be. This was all documented before operation. After operation, the following details were recorded: laryngoscopy grade (after intubation in the case of fibreoptic intubations), other factors (i.e. not dental abscess related) contributing to difficult intubation, and inter-incisor distance measurement after intubation. A free-text section allowed colleagues to comment on the system and the airway management outcomes.
|
|
The data were recorded on a spreadsheet (Excel) while the statistics used were mainly descriptive with
2 trend analysis for ordinal data.6 |
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Thirty-two patients were recruited. Their ages ranged from 18 to 69 yr (mean 35 yr, SD 14) and 17 were male. In only three cases, did anaesthetists consider patients as having features other than dental abscess that might cause difficult intubation. In two, laryngoscopy was normal anyway, and in the third, the grade 3 laryngoscopy was considered more likely to have been caused by the dental abscess. Anaesthetists recorded risk factors for abscess rupture (Part B) and some or all of these signs/symptoms were present in 10/15 cases of floor of mouth only infections. Rupture risk score
3 occurred in nine, of whom eight had floor of mouth involvement. Despite this, there were no definite ruptures in any of these cases at any time. Problem/management matrix form completion was reasonably consistent. Twenty-four of 32 cases had awake fibreoptic intubation, and this was prompted as the obvious choice in 19 by the matrix mainly on the basis of trismus. Of the other five, this was a reasonable choice in two, user preference in two others, and incorrect matrix summation in one. In the remaining 8/32, direct laryngoscopy was chosen because the matrix summation allowed alternatives. The basic plan, that is fibreoptic or direct laryngoscopy, was then interpreted for the individual patient with the appropriate contingency table (Tables 3 and 4). Overall, completing the contingency tables was very good in 23/32, incomplete in seven, and not completed at all in two.
For conventional Macintosh laryngoscopy, we describe five steps. The responses for the left columns were surprisingly varied. For Step 1, the comments indicated a degree of confusion in the sense that limited mouth opening was meant to indicate worsening after induction of anaesthesia (e.g. masseter spasm) rather than the initial trismus due to dental abscess. In Step 2, there was universal agreement that LMA should be used. In contrast, Step 3 showed markedly diverse solutions, yet only one anaesthetist referred to the BURP7 manoeuvre. For Step 4, the use of smaller tracheal tubes was documented by some but not all anaesthetists. For the right-hand side of the contingency tables, the relevance of the ‘things make it worse’ depended on what problems were outlined at the preoperative visit so only general comments are represented. Overall, these appeared to be the more challenging parts to complete and solutions appeared well thought out, relevant, and feasible.
For the left-hand side of the awake fibreoptic contingency table, Step 1 gave rise to quite diverse answers all of which were well thought out. Expected responses were given at Step 2, except that most anaesthetists only listed one of many alternative solutions. For Step 3, the majority response was to check for correct tube placement with the fibrescope rather than a CO2 trace. In Step 4, most anaesthetists opted to remove the whole unit in response to difficulty in removing the fibrescope from within the tube. For loss of i.v. access (Step 5), most anaesthetists suggested that they would use inhalation induction with sevoflurane rather than attempt to re-site the i.v. cannula. Again the right-hand side of the fibreoptic table responses were more patient specific and appeared more challenging to complete. Solutions again appeared well thought out, relevant, and reasonable. It was noteworthy that there was a high likelihood of early progress to tracheostomy under local anaesthesia in the event of a worsening situation. Unfortunately, in the four cases where tracheal access was scored as abnormal (i.e. >1/5) in the preoperative assessment, this did not appear to be reflected appropriately in the contingency table completion. Only one form used the words ‘consider’ trachesotomy rather than implying it would be straightforward.
Three cases are worthy of note and were managed by senior anaesthetists. A 51-yr-old male had a pharyngeal abscess and was scored 4 for toxicity, trismus (inter-incisor distance 0.5 cm), and distortion. At Stage 1 of the fibreoptic management, difficult anatomy and prolonged intubation time led to desaturation. Remifentanil was being used for sedation, so this was reversed and the procedure interrupted. It was restarted soon afterwards and intubation completed without event. Another 22-yr-old male with a floor of mouth/masticator abscess had scores of 4/5 for trismus (inter-incisor distance 1.5 cm) and 5/5 for risk of rupture. At Stage 3 in fibreoptic management, the tube could not be confirmed as positioned in the trachea. It was assumed that it had become displaced and curled up in the pharynx. After withdrawing the tube slightly, the fibrescope was repositioned at its tip, so that both could go through the laryngeal inlet together, effectively intubating under direct vision. The third case was a 20-yr-old male with an abscess affecting superficial, floor of mouth, and masticator spaces. He had maximum scores (5/5) for toxicity, trismus (inter-incisor distance 0.5 cm), distortion, and risk of rupture. Fibreoptic management continued to Stage 5 without event when his compliance (initially rated excellent at 1/5) became an issue. He attempted tracheal tube removal but as this was anticipated, he was immediately given an i.v. induction of anaesthesia and intubation was completed without further event.
The preoperative clinical assessments as to which spaces were involved compared well with the intraoperative surgical findings and in only 6/32 cases did they not match. Differences were considered trivial in three (i.e. which particular floor of mouth spaces were actually involved). Another case assessed as a floor of mouth infection, turned out to only involve the buccal space. (Effectively, this could be considered as an overstatement of the implications for laryngoscopy.) In two others, the assessment understated the clinical significance. In the first, a masseteric space infection was diagnosed before operation, but at operation both pharyngeal and floor of mouth spaces were also involved. The final one assessed as superficial turned out to involve the floor of mouth instead. In both these cases, the error could have led to unexpected difficulty with laryngoscopy.
The cervical fascial spaces involved in these cases and the laryngoscopy grades8 are described in Table 5. Floor of mouth involvement, alone or in combination, affected laryngoscopy grade. 2 trend analysis for laryngoscopy grade vs all other space involvements was 4.57 (P<0.05). Table 6 shows the interaction between difficulty (i.e. grades 3 and 4), floor of mouth involvement, and trismus. If the floor of mouth was not involved, the risk of difficulty was 1/13 (8%). However, if floor of mouth was involved, this rose to 6/15 (40%, P<0.05). Furthermore, 4/9 remaining cases did not have laryngoscopy grade recorded after fibreoptic intubation (and all of these had trismus that failed to improve after intubation) so this is almost certainly an under-representation.
|
|
Figure 4 shows the relationship between inter-incisor distance (pre- and post-induction) and superficial only vs other space involvement. The preoperative values for inter-incisor distance improved after induction to varying degrees for both and we found no predictor of improvement for individual cases. The one case of pharyngeal abscess started with an inter-incisor distance of 0.5 cm which did not improve after induction. Otherwise, it is worth noting that no case of superficial space only involvement had an inter-incisor distance of 2 cm or less after induction. For other spaces, if the preoperative inter-incisor distance was 2 cm or less, there was only a 50% chance that it would be >2 cm after induction.
|
As far as clinical significance of space involvement was concerned, the following points are worthy of note (relative to Table 1). As noted above, tracheal access was highlighted as likely to be a problem in four and all of these had floor of mouth infection. There was only one case of pharyngeal spread and this case was grade 3 laryngoscopy thought to be due to reduced oro-pharyngeal space and severe trismus which did not improve after induction. Masticator space involvement is usually regarded by surgeons as having potential for severe trismus which may not improve after induction.9 This was not so for 3/4 of the cases where it was involved in this series. There were no cases of distant spread of infection or Ludwig's angina in this series.
Eight out of nine cases with floor of mouth involvement (alone or in combination) and assessment scores of more than 15 were all anaesthetized by consultants. This effectively meant all patients who had moderate toxicity, airway distortion, risk of abscess rupture, and trismus. In contrast, however, there were three cases recorded as grade 4 laryngoscopy anaesthetized by trainees. All had floor of mouth infections (alone or in combination) and fibreoptic intubation had been undertaken because of trismus. (The laryngoscopies were performed post-intubation.) None was assessed as having obvious severe airway distortion or toxicity when assessed before operation.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
This study required preliminary training for the new system, review of the relevant anatomy, and getting colleagues enthusiastic enough to engage in seeking consent and form filling. It is therefore inevitable that this will have influenced their efforts in using the system and the results we obtained. Indeed, this enthusiasm was transmitted to surgical colleagues who were peripherally involved in confirming which spaces were infected. At this stage, the system should be seen as relatively raw in the sense that it was important to understand how much variation occurred in the clinical acumen of both trainees and seniors. This was particularly impressive because we were dealing with elective cases and it is reasonable to assume that inconsistency would be more of a problem in urgent situations. Another limitation might be seen to be the small number of management options that were considered. However, this was never mentioned as an issue by any of the anaesthetists and so, as expected, was representative of the local practice. It is our opinion that this system is easily adapted for other difficult intubation scenarios and other management options.
Proving the efficacy of the new system with so few cases is obviously impossible. However, we have defined some of the problem areas as far as consistency of approach is concerned when otherwise this might have been taken for granted (i.e. the notion that everyone would always ‘do the right thing’). This provides enormous potential for teaching and investigation as to why the discrepancies. Nevertheless, we were reassured that in no case was the basic plan abandoned and no events occurred that had not been anticipated by the planning system. The three cases described relevant to the appropriate steps and contingency table seemed to imply that the anaesthetists concerned had anticipated the problems outlined having had the opportunity to review the ‘worst case’ options immediately before operation. This was in sharp contrast to our experience in the earlier audit where unexpected events lead to rushed and less acceptable managements. Highlighting factors known to be associated with abscess rupture may also have been relevant to this not being seen in any case, despite the moderate overall risk demonstrated. The discrepancy between preoperative assessment and contingency planning was probably an issue as far as notes about tracheal access were concerned. This would imply an advantage to computerization of the process, so that any tendency to overlook information already documented could be flagged up. Generally, feedback (response to the free-text comments section) from anaesthetists using the system was positive and constructive, despite their clear expressions of concern about the time it took to complete the documentation.
Currently, there are no UK guidelines for the approach to the anticipated difficult airway. It has been our view for some time that a scheme based on plan A with a reserve plan B10 is not the best approach. Binary plans encourage a binary approach when many reported problems can rather be considered as a serial worsening.11–14 Our approach is based on the idea that things will go wrong and this should be planned for rather than assuming it will never happen. ‘On the hoof’ planning (as occurred in our earlier audit) must be considered a recipe for disaster. Complicated airway problems such as Ludwig's angina pose management issues at different levels on the way to successful intubation and only by being clear about the implication at all stages can planning be considered to have been adequate. The plan A/plan B philosophy often means using a different route for intubation (e.g. awake fibreoptic nasal intubation first and if that fails, try general anaesthesia and conventional direct laryngoscopy). When these have markedly different risks, there is the possibility that the second option will be started under adverse circumstances (e.g. when nasal haemorrhage has already occurred). In addition, it is known that repeated attempts at various airway manoeuvres tend to make matters worse.15 (There will always be a temptation to persist with the safer option when anaesthetists do not want to progress to the less safe alternative.)
In many situations, alternative intubation routes will not be available. In that case, everything must be done so as not to abandon the only route available. The answer to our mind is appropriate structured planning based on ‘worst case’ and making the risk of failure negligible. For example, if only the nose is available, immediately before induction is not the time to be checking its patency. For any given clinical scenario, it should be possible to construct a ‘problem list’ based on literature review or expert consensus. This then acts as a prompt list for the anaesthetist seeing the patient before operation. The problem/management matrix is constructed from the patient-specific problem list. Careful review of the relevant contingency table should highlight avoidable risks, suggest where the anaesthetist's competence would be inadequate and situations that should be avoided at all cost. All of this needs to be computerized to be clinically useful and this in turn should mean that cases could be documented, reviewed, and the database itself revised in the light of improved knowledge. Although the system cannot hope to provide infinite solutions, it is useful in encouraging a logical approach to anticipated difficult airway problems.
Anaesthetists should be rightly concerned about difficulty in judging floor of mouth rigidity and airway distortion in the face of moderate trismus. When the floor of mouth is involved, extra vigilance is appropriate in assessing these patients. This becomes even more critical when tracheal access is also compromised. As far as Responsive Contingency Planning is concerned, these observations have two important implications. First, floor of mouth rigidity was not included originally and this needs to be reflected by updating the problem list. Secondly, we were concerned that occasional inconsistencies occurred when anaesthetists suggested awake tracheostomy as an option when tracheal access had already been highlighted as difficult. This again stresses the use of the new system, even as a paper exercise, in teaching a structured approach to this sort of complex group of clinical problems.
|
Acknowledgements |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Anaesthetic colleagues were important for this study because they were responsible for gaining consent for their own patients’ involvement and also contributed by filling in forms and reports. Their opinions were invaluable in assessing the new system and its implementation and we would like to express our sincere gratitude for their co-operation.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
1 Perry M. Infections. In: Head, Neck and Dental Emergencies—Perry M, ed. (2005) Oxford: Oxford University Press. 69.
2 Green AW, Flower EA, New NE. Mortality associated with odontogenic infection. Br Dent J (2001) 190:529–30.[CrossRef][Web of Science][Medline]
3 Bennett JD, Flynn TR. Anesthetic considerations in orofacial infections. In: Oral and Maxillofacial Infections—Topazian RG, Goldberg MH, Hupp JR, eds. (2002) 4 Edn. Philadelphia: W B Saunders. 439–55.
4 Neff SPW, Merry AF, Anderson BJ. Airway management in Ludwig's angina. Anaesth Intensive Care (1999) 27:659–61.[Web of Science][Medline]
5 Scott JH. The spread of dental infection—anatomical considerations. Br Dent J (1952) 38:352–8.
6 Altman D. Comparing groups—categorical data. In: Practical Statistics for Medical Research—Altman D, ed. (1991) 1st Edn. London: Chapman and Hall. 229–76.
7 Knill RL. Difficult laryngoscopy made easy with a ‘BURP. Can J Anaesth (1993) 40:4.[Web of Science][Medline]
8 Cormack RS, Lehane J. Difficult tracheal intubation in obstetrics. Anaesthesia (1984) 39:1105–11.[Web of Science][Medline]
9 Balatsouras DG, Kloutsos GM, Protopapas D, et al. Submasseteric abscess. J Laryngol Otol (2001) 115:68–70.[CrossRef][Web of Science][Medline]
10 Mason RA, Fielder CP. The obstructed airway in head and neck surgery. Anaesthesia (1999) 54:625–8.[CrossRef][Web of Science][Medline]
11 Shaw IC, Welchew EA, Harrison BJ, Michael S. Complete airway obstruction during awake fibreoptic intubation. Anaesthesia (1997) 52:582–5.[CrossRef][Web of Science][Medline]
12 Ho AMH, Chung DC, To EWH, et al. Total airway obstruction during local anesthesia in a non-sedated patient with a compromised airway. Can J Anaesth (2004) 51:838–41.[Web of Science][Medline]
13 Piepho T, Thierbach A, Werner C. Nasotracheal intubation: look before you leap. Br J Anaesth (2005) 94:859–60.
14 Mort TC. Emergency tracheal intubation: complications associated with repeated laryngoscopic attempts. Anesth Analg (2004) 99:607–13.
15 Burkle CM, Walsh MT, Harrison BA, Curry TB, Rose SH. Airway management after failure to intubate by direct laryngoscopy: outcomes in a large teaching hospital. Can J Anaesth (2005) 52:634–40.[Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||